In the past emotion had very often been formulated to be a transient, internal and intra-personal phenomena with behavioral, physiological and subjective components. This view is being challenged by many researchers (e.g., Abu-Lughod & Lutz, 1990; Lutz & White, 1986; Parkinson, 1996; Rimé, Mesquita, Philippot, & Boca, 1991) and the research on social sharing and mental rumination associated with emotional experiences clearly shows that emotion is neither a strictly personal nor a short lived phenomenon.
Social sharing of emotion is the term used to describe the process during which a person having experienced an emotion recounts this experience to his/her social environment. This essentially entails the transmission of information and experience of a personal and emotional nature from the person experiencing the emotion to his or her socio-cultural environment. The social sharing literature has clearly established that over 80% of our emotional experiences are shared (Rimé et al., 1991; Rimé, Philippot, Boca, & Mesquita, 1992; Rimé, Finkenauer, Luminet, Zech, & Philippot, 1998). In most cases the sharing of an emotion after an experience is immediate and frequent.
Mental ruminations have also been widely associated with emotional experiences. Mental rumination has been defined as «conscious thought directed towards a given object for an extended period of time» (Martin & Tesser, 1989). These authors see ruminations as involving attempts to reach important unattained goals or reconciling oneself to not reaching those goals. Ruminations have also been defined as thoughts, mental images, or memories related to an event or to the person/people associated with it (Tait & Silver, 1989). A similar approach is to see ruminations as «passively and repeatedly focusing on one’s symptoms of distress and the circumstances surrounding those symptoms» (Nolen-Hoeksema, McBride, & Larson, 1997). Excessive rumination has been linked to prolongation of the distress caused by an event and may be a stable characteristic of some people (Nolen-Hoeksema et al., & Larson, 1997; Nolen-Hoeksema, Parker, & Larson, 1994).
Although there is a wide acceptance of the idea that the emotional lives of men and women are different from each other, little is known about the differences between the way in which men and women share and ruminate about their emotions. Even less is known about whether men and women differ from each other in similar ways in different cultures. There is almost a universal stereotype concerning female emotionality, women are supposed to express and experience more emotion. The research on display rules and defence mechanisms shows gender differences, girls tend to inhibit negative affect while turning them against themselves and boys tend to neutralise the expression of most kinds of affect and they project negative feelings externally (Brody, 1985). The self-report data appears to corroborate this idea of female emotionality. There is a growing belief that the stereotype that females are more emotional than males is based on the deficit model, a belief that males do not express the emotion they feel (Fabes & Martin, 1991).
The various theoretical explanations of gender differences stress differences in status or social roles of men and women. Gender is seen to be a component of an ongoing interaction where one person in an interaction emits expectancies and the self negotiates its identity, the context of this interaction between the self and another person shapes the behavior in question (Deaux & Major, 1987). Thus it would imply that the expression of gender takes place in the context of social interaction and would be manifested in the personality and emotional processes.
The process of socialisation is generally thought to be at the heart of any differences that may exist between males and females. It is essentially a process by which one learns to behave in a socially appropriate way, to play the appropriate social roles and develop social attitudes and values. The socialisation process is most active through childhood and adolescence. Sex differences in the socialisation process leads to differences in behaviour of the two sexes. In general terms girls are socialised more towards nurturance, responsibility and obedience and boys more towards independence, self-reliance and achievement (Berry, Poortinga, Segall, and Dasen, 1992).
Sex differences in personality development have been found to arise in late childhood and gradually decline in the college sample (Cohn, 1991). These sex differences could be attributed to differences in biological maturation or to differences in socialisation experiences. Although there is debate on whether there are real differences between males and females on personality measures, it is accepted that gender differences do exist especially when one is looking at social behaviour (e.g., Maccoby, 1990)
Social support research suggests that women report greater perceived social support, greater satisfaction with social support than men and they also record higher on socially skilled responses in an interaction (Sarason, Sarason, Hacker, & Basham, 1985). Women have been found to experience emotions associated with powerlessness, such as fear, sadness and shame, more intensely and more frequently (Fischer, 1994). No differences between the emotionality of men and women where emotions related to feeling powerful are concerned were found. One explanation offered by this research is that men view emotions associated with powerlessness more negatively than women do and it induces them not to experience these emotions often or intensely.
The issue of emotional intensity has been linked widely to gender differences. Female respondents report experiencing a higher intensity of emotional experience (Diener, Sandvik, & Larsen, 1985; Larsen & Diener, 1987) and it appears that the sex differences in emotional intensity derive from sex-differentiated normative pressures that specify that women are more emotionally responsive than men (Grossman & Wood, 1993). The gender differences may be linked to the cultural expectation for women to be more expressive and amplify their emotional responses.
The diversity in the emotional life of men and women has been observed in western European and North American samples. In these samples there is also considerable evidence to suggest that boys and girls prefer to, and do interact in increasingly gender segregated groups. The predominant features of boy groups is competition, dominance, rough and tumble play; girl groups on the other hand are characterised by close and intimate friendships and a facilitative interaction style (Maccoby, 1990). The behaviour patterns formed in early peer interactions lead to the kind of differences in adult behaviour mentioned above.
There is some evidence to suggest that the cultural and gender differences in self-construal can not be explained by the same dimensions (Cross & Madson, 1997; Kashima, Yamaguchi, Kim, Choi, Gelfand, & Yuki, 1995). The «relational» dimension has been proposed to account for the gender differences in psychological functioning (Gilligan, 1982; Kashima et al., 1995). This dimension represents the extent to which people see themselves as emotionally related to others and is independent of the individualism-collectivism dimension commonly used to account for cultural differences. While some studies argue for universality of gender differences in self construal (Miller, 1994), others have reported stronger male stereotypes in countries with a higher power distance and lower levels of literacy and socioeconomic development (Williams & Best, 1990).
This paper presents some results from two studies conducted on adolescents in order to examine the social sharing and mental rumination patterns in adolescence. The first study examined the social sharing of fear, shame and sadness and the second study compared shared events with events held secret on dimensions of emotional quality and mental rumination. Adolescents came from three cultures, namely, Indian, immigrant Indian, and English. These cultures represent different levels of individualism, the English culture is rated as being highly individualistic and the Indian culture the least individualistic (Hofstede, 1980). The immigrant Indian group living in England forms a true ethnocultural group (Berry et al., 1992) and resides in a culture very different from its culture of origin.
Only the results relating to the variable sex will be addressed in this paper as the rest of the results have been extensively analysed elsewhere (Singh-Manoux & Finkenauer, 1999a, 1999b). In general terms the individualism and collectivism dimensions were not expected to be able to explain the difference between the two sexes as it appears highly unlikely that the same dimensions of psychological functioning underline both gender and cultural differences. Girls were predicted to register a higher rate and frequency of sharing in keeping with the findings of higher female emotionality. The deficit hypothesis which suggests that males do not express the emotions they feel led to the prediction that female participants would initiate sharing more and also would share feelings rather than superficial details of the emotional situation.
In the second study female subjects in all cultures were expected to register higher intensity and distress associated with their emotional experiences. As there is some evidence to suggest that females are more likely to report ruminations (Nolen-Hoeksema et al., 1994; Rimé et al., 1992), the female adolescents were expected to report more frequent ruminations. Both studies were exploratory and intended to examine the sex differences in different cultures.
The participants were 9th and 11th graders from public schools in New Delhi in India and London in the United Kingdom. Participation of the respondents was solicited through the help of the school authorities and no payment was made either to the schools or to the respondents. In all 585 adolescents responded to the questionnaire; 30 (5.1% of all questionnaires) of these were rejected on grounds of being incomplete. The final sample was composed of 179 Indian adolescents from New Delhi (94 male and 85 female), 182 immigrant Indian (89 male and 93 female) and 194 English adolescents from London (99 male and 95 female). The average age of the 9th graders was 14.71 (SD = 0.46) and that of the 11th graders was 16.82 (SD = 0.39). 51.5% of all respondents were from the 9th grade and 48.5% of them were from the 11th grade; and there was no significant difference in the number of 9th and 11th graders in the three cultural groups (2(2, N = 555) = 1.15, n.s.).
The respondents were recruited through high schools in the middle-class suburbs of the cities targeted (London in England and New Delhi in India) by this study. Although it is always difficult to ensure equivalence across samples, the socio-economic status and the educational level of the schools represented in the sample was very similar. All the participating schools drew their students from the urban middle-class and had a homogeneous population. All the immigrant Indian adolescents who participated in this research were British citizens and were born and brought up in England.
All participants responded to a questionnaire in english as this was the language of instruction in the participating schools. English is one of the national languages of India. Hindi, the other national language, is mostly employed in the north of India and as the present sample was not composed exclusively of Hindi speakers, the English version of the questionnaire was the practical choice for the Indian sample.
The construction of the questionnaire was based on semi-structured, exploratory interviews carried out on adolescents in India and England. In addition, some of the items were drawn from the various questionnaires utilised by Rimé et al. (1991, 1992) in their investigation into social sharing. The final version of the questionnaire was pilot tested to ensure that participants from the target age groups understood the questionnaire and were capable of responding to it. The participants were asked to briefly describe the situation which led to the experience of the emotion (fear, shame and sadness), and were then asked to respond to the questions.
The rate of sharing was assessed simply by asking the participants whether or not the emotion had been shared. Subsequently, they answered questions relating to the delay of sharing (never, same day, same week, same month or later), the frequency of sharing (never, only once, 2 to 4 times, 5 times or more) and the number of people with whom the experience had been shared (nobody, only one person, 2 to 4 people, 5 people or more).
This was followed by questions concerning the identity of the sharing partners (parents/grandparents/uncle/aunt; brother/sister/cousin; stranger; a professional; boy/girl friend; friend). On three point scales (1 = never; 3 = always), the respondents were then asked whether the sharing had been initiated by themselves, the sharing partner or external factors.
Subsequently, the content of the sharing (asked for advice/help; spoke about what had happened in detail; spoke about what you had felt) was assessed. Finally, the perceived effects of sharing by the participant were assessed on 5-point scales (1 = not at all, 5 = very much): (1) feeling better, (2) feeling relieved, (3) feeling no different, (4) feeling closer to the person with whom shared, (5) feeling calmer, and (6) feeling worse.
Rate, delay, and frequency of sharing
Although there was no main effect for sex on the rate of sharing (partial G2 (1) = 0.72, n.s.), the delay of sharing (F (1, 549) = 2.74, n.s.), the number of times an emotion was shared (F (1, 549) = 0.05, n.s.) and the number of people with whom it was shared (F (1, 549) = 2.27, n.s.) there were sex x culture interactions. The interaction for the rate of sharing (partial G2 (2) = 7.02, p<.05) revealed that English female respondents reported a higher sharing rate as compared to those from India (see Table 1). The sex and culture interaction affected the frequency of sharing both in terms of number of times of sharing (F (2, 549) = 5.23, p<.01) and the number of people with whom shared (F (2, 549) = 4.54, p<.01). For both these measures of frequency the results, as is clear from Table 1, indicate that the Indian males showed more frequent sharing and it is the English females who showed more frequent sharing. The immigrant Indian male and female adolescents did not differ from each other on measures of sharing rate or frequency.
Identity of sharing partner
In order to examine the effect of sex and its interaction with culture on the choice of sharing partner, a log-linear analysis was carried out. There was no evidence of an interaction between sex and culture (partial G2 (10) = 15.01, n.s.) or for a main effect for sex (partial G2 (5) = 1.93, n.s.). It is mostly intimates who were chosen as sharing partners both by male and female adolescents. Friends formed the preferred sharing partners for both males (50.4%) and females (50.5%), followed by parents (for 25.8% of males and 24.5% of females) and siblings (for 16.4% of males and 16.9% of females).
The male and female adolescents only differed in the extent to which they themselves reported initiating the sharing process (F (1, 549) = 9.15, p<.01), with females in all three cultures reporting initiating the sharing themselves to a greater extent. There were no differences between the sexes when the triggering of sharing was attributed to the sharing partner or to another external source.
Content of sharing
The variable sex affected the content of sharing (partial G2 (2) = 17.91, p<.001) in the direction predicted. As can be seen from the Table 2, as compared to the female participants a higher proportion of the male adolescents reported superficial sharing and a lower proportion reported sharing their feelings. The lack of sex and culture interaction (partial G2 (4) = 5.09, n.s.) revealed that this pattern of results is held in all the three cultures examined.
Perceived effects of sharing
A culture (Indian vs. immigrant Indian vs. English) x sex (male vs. female) x effects of sharing (feeling better, relieved, no different, closer to the sharing partner, calmer, worse) MANOVA revealed a multivariate effect for sex (F (6, 250) = 2.93, p<.01). An examination of the univariate results revealed that this effect was accounted for by the female participants reporting feeling closer to the sharing partner (M = 2.93) as a result of sharing to a greater extent as compared to the male participants (M = 2.46).
In all 268 adolescents responded to the questionnaire, 17 questionnaires (7%) were rejected on grounds of being incomplete. The final sample was composed of 251 adolescents with a mean age of 16.85 (SD = 0.38). Eighty-six of these adolescents were from New Delhi in India (48 male and 38 female), 84 were immigrant Indian adolescents from London (42 male and 42 female), and 81 were English adolescents from London (40 male and 41 female).
Here again all adolescents were recruited from high schools located in the suburbs of a big city (New Delhi in India and London in Britain). All adolescents followed normal schooling and were in the 11th grade at the time they participated in this study. The immigrant Indian participants were born in Britain, but were bilingual, speaking English and Hindi or Punjabi fluently. The questionnaire was administered in English in all the three cultures, for reasons similar to the ones stated in study 1.
The questionnaire was adapted from Finkenauer and Rimé (1998). It consisted of two identical sets of questions with the first set referring to a shared emotional event, the other referred to a secret event. Participants were first requested to remember a secret (shared) emotional event, before answering the questions. Participants rated on a 7-point scale (1 = very weak; 7 = very strong) the emotional intensity of the event. Then they were asked to rate the distress caused by the event on an 11-point scale (1 = no distress; 11 = extremely distressful). These ratings were requested twice, once for the time the event occurred and once for the present, when remembering the event.
The questionnaire comprised of ten appraisal dimensions commonly considered in the literature: Valence, Moral Evaluation, Self-Evaluation, Responsibility, Control, Certainty, Familiarity, Expectedness, Agency, and Importance. Four sets of antonyms assessed each dimension (e.g., pleasant-unpleasant, desirable-undesirable, positive-negative, a success-a failure, for Valence; strong-weak, in control-without control, efficient-inefficient, powerful-helpless, for Control). Participants rated each set of antonyms on 7-point bipolar scales with the antonyms as anchors (e.g., 1 = pleasant; 4 = neither / nor; 7 = unpleasant). For each dimension, the four items were averaged to obtain individual scores.
Then nine items on 7-point rating scales assessed the belief changes brought on by the experience. The impact of the event on the individual was judged through a set of 14 items drawn from Post Traumatic Stress Disorder (PTSD) checklist which were rated by the subject on 7-point scales. For both these measures the items were averaged to obtain individual scores.
This was followed by six measures (expression of feeling, behavioural expression, dissimulation of feeling, apathy, feelings of paralysis, loss of control) of emotion regulation behaviours on 7-point scales. Subsequently the frequency, duration, intensity, and intrusiveness of mental ruminations were assessed on 5-point scales. Finally, participants rated on 7-point scales (1 = not at all; 7 = very much) different types of mental rumination associated with the event: (1) putting order in what happened, (2) remembering all the details, (3) mentally modifying what has happened, (4) understanding what happened, (5) imagining how things would be if the event had not occurred, and (6) finding meaning in what happened.
Intensity and distress of events
The intensity (F (2, 245) = 7.44, p<.001), initial distress (F (2, 245) = 3.88, p<.05) and current distress (F (2, 245) = 7.82, p<.001) of reported events was found to involve an interaction between culture and sex. As can be seen from Table 3, male adolescents reported more intense events and more distress in the Indian culture whereas it was the female adolescents who reported higher intensity and distress in the English culture.
A culture (Indian vs. immigrant Indian vs. English) x sex (male vs. female) x cognitive appraisal (10 appraisal dimensions) MANOVA yielded a culture x sex interaction (F (20, 472) = 2.42, p<.001) and a main effect for sex (F (10, 236) = 3.19, p<.001). The interaction between culture and sex was explained by the female participants in the English culture appraising the event as less controllable and the self as more blameworthy. The differences between the two sexes, although not significant, were in the opposite direction in the collectivistic cultures. The main effect for sex was accounted for by female respondents in all cultures reporting their emotional experiences to be more negatively valenced and as being immoral to a greater extent when compared to males.
Belief changes and trauma symptoms
Female subjects in this sample reported more belief changes than the male subjects (F (1, 245) = 11.77, p<.001) in all three cultures examined, suggesting that emotional events had more of an impact on the assumptions that female adolescents hold about themselves and the world in general when compared to male adolescents. Analysis revealed a culture x sex interaction on the measure of trauma caused by the emotional event (F (2, 245) = 8.04, p<.001). This interaction was explained by the female adolescents reporting more intense symptoms of trauma in the English and Immigrant Indian sample but not in the Indian sample.
Emotion regulation behaviour
A culture (Indian vs. immigrant Indian vs. English) x sex (male vs. female) x emotion regulation behaviour (expression of feeling, behavioural expression, dissimulation of feeling, apathy, feelings of paralysis, loss of control) MANOVA revealed a main effect for sex (F (6, 240) = 4.74, p<.001). The effect for sex was accounted for by two univariate effects: female adolescents reported expressing their emotion and feeling a loss of control to a greater extent as compared to the male adolescents. There was no interaction between culture and sex on the dimensions of emotion regulation.
Mental rumination variables
As is clear from Table 4, the female participants in all cultures reported a higher frequency of mental rumination (F (1, 245) = 6.90, p<.01). The duration, intensity, and intrusiveness of mental rumination was not found to be influenced by sex as the male and female scores did not differ from each other on these measures (see Table 4). A culture (Indian vs. immigrant Indian vs. English) x sex (male vs. female) x mental effort (6 types of mental efforts) MANOVA was conducted to examine the effects for sex. The analysis revealed a main effect for sex (F (6, 93) = 3.93, p<.01). As shown in Table 4, this effect was due to the male adolescents reporting more efforts to put order in the event and attempting to modify the course of things to a greater extent.
The main goal of this paper was to examine social sharing and mental rumination linked to emotional experiences in adolescence with particular attention to the differences between the two sexes. Consistent with previous research, the findings from the two studies reported here did show differences between the two sexes. The unexpected result is that the stereotype of females as being more emotional holds true more in the English culture as compared to the Indian culture. The sex differences in the immigrant Indian group are weaker than those in the other two groups.
The first study provides some evidence for the belief that the emotional lives of men and women differ in important ways. The stereotype of stronger female emotionality holds true particularly in the individualistic, developed country examined in this study. The English female participants show a higher rate of sharing and a higher frequency of sharing. Female participants in all cultures report initiating the sharing themselves to a greater extent showing greater skill and ease with which they communicate emotional information. This is also reflected in their sharing their feelings rather than superficial details relating to the emotional event. The male subjects report sharing situational details rather than their feelings. Finally, female subjects appear to perceive the relational benefits of sharing to a greater extent by reporting that sharing led to greater feelings of closeness with the sharing partner. The more intimate sharing among females reflects the findings in the domain of social support (cf. Sarason et al., 1985).
The difference between the emotional lives of female and male adolescents retained in the first study was further corroborated in the second study. It had been hypothesised that female subjects would report higher intensity and distress related to the emotional events. This hypothesis was only partially supported as it is only the English female participants who reported more intense and distressful events. In fact in the Indian culture it is the male participants who reported more intense and distressful events.
The manner in which events were appraised by the two sexes also differed with female subjects in all cultures reporting events to be more negatively valenced and immoral. However, some differences between the two sexes existed only in the English culture. Only in the English culture did the female adolescents appraise the emotional events as being less controllable and perceived the self as more blameworthy.
The emotional impact of the situation was measured through the belief change and the symptoms of trauma engendered by the event. One of the consequences of an emotion is the temporary belief change that it brings about (Frijda, Mesquita, Sonnemans, & Van Goozen, 1991). The belief change involves the individual generalising aspects of a particular emotional situation or its consequences to other situations. The results here suggest that the female subjects in all cultures report greater belief changes as a result of having experienced an emotion.
It seems that for negative life events, the general stress response tendency although less extreme, is similar in many respects to the intrusive elements associated with Post Traumatic Stress Disorder (Tait & Silver, 1989). Negative life events have been linked to clinical psychological symptoms by many researchers. The relationship between negative life events and psychological symptoms appears to be pretty robust as response bias alone cannot account for this relationship (Lakey & Heller, 1985). In this study the English and the immigrant Indian female adolescents reported more intense symptoms than male adolescents in these cultures. There were no sex differences in the Indian culture.
The emotion regulation behaviours measure the extent to which an individual engages in behavioural expression or inhibition while experiencing an emotion. The female subjects in all cultures expressed their emotion to a greater extent and feared loosing control to a greater extent. The results suggest that male adolescents are less emotionally expressive than female adolescents.
Consistent with earlier research (Nolen-Hoeksema et al., 1994; Rimé et al., 1992) female subjects reported a higher frequency of mental ruminations. Mental ruminations related to a emotional event are a measure of the continuing involvement of the individual with the event. The results would then suggest that female adolescents continue to be more preoccupied with their emotional experiences for a longer period of time. The male adolescents in all cultures report more mental efforts at dealing with their emotions. Specifically, they engage in more attempts to put order in the event and modify the course of things. This is a more active approach to dealing with emotions as compared to the frequent mental ruminations that appear to be more characteristic of female adolescents. This difference in the way males and females respond to the consequences of an emotion may be an important dimension in the difference between the two sexes where their emotional response is concerned.
Whether the difference between male and female adolescents implies that females are «more emotional» really amounts to asking what the term «more emotional» implies. What the result obtained in these two studies clearly show is that the reported emotional experience of male and female adolescents is clearly different from each other. The deficit hypothesis, which proposes that differences in male-female emotionality can be traced to males not expressing the emotions they feel (Fabes & Martin, 1991), cannot account for all the findings. The «relational» dimension (Kashima et al., 1995) to explain the gender differences in psychological functioning refers to the extent to which people see themselves as emotionally related to others, is better able to explain the findings of this study.
An interesting feature of the observed differences between male and female adolescents is that many of these differences are observed only in the individualistic English culture. The culture which is characterised by a higher power distance and lower literacy (cf. Williams & Best, 1990) is the Indian culture and the results certainly do not support the idea of a stronger male stereotype in India. The observed interaction between culture and sex in the two studies brings indirect evidence for the hypothesis that cultural and gender differences cannot be explained by the same psychological dimensions (Kashima et al, 1995). The gender differences can perhaps be best explained by the female subjects sampling the «relational» dimension of their selves more frequently. The individualism-collectivism dimension certainly does not explain the gender differences, consequently the assumption that females in individualistic cultures are a little like the collectivists is unfounded.
The differences in the socialisation process of males and females have been found to be one of the principal sources of sex differences (Berry, et al., 1992). The results of this research indicating fewer gender differences in the Indian sample was unexpected, and can be explained either by lower or later sex differentiation in the Indian culture. The hypothesis of later sex differentiation may be more pertinent due to the way adolescence is conceptualised in the Indian culture. Adolescents are seen to be older children rather than young adults in middle class India and is likely that gender differences appear a little later in the Indian culture.
It is difficult to say whether similar sex differences will be found in adult populations as there is some evidence to suggest that sex differences are at their greatest in adolescence (Cohn, 1991). There is also evidence to suggest that the sharing and mental rumination patterns of male and female adults from western European samples does not differ very much (Rimé et al, 1991, 1992). Therefore, an important future task would be to investigate the sex differences in adults and adolescent samples more systematically. It would also be important to examine these differences from a cross-cultural perspective in order to assess if these differences exist in adolescence alone.